REAL ESTATE VALUATION THEORY

RESEARCH ISSUES IN REAL ESTATE

Sponsored by the AMERICAN REAL ESTATE SOCIETY

Volume I

APPRAISAL, MARKET ANALYSIS, AND PUBLIC POLICY IN REAL ESTATE

edited by James R. DeLisle and J. Sa-Aadu

Volume II

ALTERNATIVE IDEAS IN REAL ESTATE INVESTMENT

edited by Arthur L. Schwartz, Jr. and Stephen D. Kapplin

Volume III

MEGATRENDS IN RETAIL REAL ESTATE edited by John D. Benjamin

Volume IV

SENIORS HOUSING edited by Michael A. Anikeef and Glenn R. Mueller

Volume V

ETHICS IN REAL ESTATE edited by Stephen E. Roulac

Volume VI

ESSAYS IN HONOR OF JAMES A. GRAASKAMP: TEN YEARS AFTER

edited by James R. DeLisle and Elaine M. Worzala

Volume VII

REAL ESTATE EDUCATION THROUGHOUT THE WORLD: PAST, PRESENT AND FUTURE

edited by Karl-Werner Schulte

REAL ESTATE VALUATION THEORY

edited by

KoWang California State University, Fullerton U.S.A.

Marvin L. Wolverton University of Nevada Las Vegas U.S.A.

1lIr....

" SPRINGER SCIENCE+BUSINESS MEDIA, LLC

Library of Congress Cataloging-in-Publication Data

Real estate valuation theory I edited by Ko Wang and Marvin L. Wolverton. p.cm. - (Research issues in real estate ; v. 8)

Includes bibliographical references. ISBN 978-1-4613-5299-0 ISBN 978-1-4615-0909-7 (eBook) DOI 10.1007/978-1-4615-0909-7 1. Real property - Valuation. 1. Wang, Ko, 1955-11. Wolverton, Marvin L. III.Series.

HD 1387 .R394 2002 333.33 '2-dc21

Copyright © 2002 bySpringer Science+Business Media New York OriginaIly published by Kluwer Academic Publishers in 2002 Softcover reprint ofthe hardcover Ist edition 2002

2002022580

AlI rights reserved. No part of this work may be reproduced, stored in a retrieval system, Of transmitted in any form or by any means, electronic, mechanical, photocopying, microfilming, recording, or otherwise, without the written permission from the Publisher, with the exception of any material supplied specifically for the purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the work.

Pennission for books published in Europe: pennissions@wkap.nl Permissions for books published in the United States of America: permissions@wkap.com

Printed on acid-free paper.

2001 American Real Estate Society

President's Council

AIG Global Real Estate Investment*

Appraisal Institute

Fannie Mae Foundation

Fidelity National Title Insurance Company

John Hancock Real Estate Investment Group

Institutional Real Estate, Inc,

LaSalle Investment Management

Regents

Cornerstone Real Estate Advisers Inc.

CoStar Group

Ferguson Partners

International Council of Shopping Centers (ICSC)

Sponsors

AEW Capital Management

Association of Foreign Investors in U.S. Real Estate (AFIRE)

Ballard, Biehl & Kaiser

BDO Seidman

BRE Properties'

CIGNA Investments

Citadel Realty

Counselors of Real Estate (CRE)

Dearborn Real Estate Education

The Dorchester Group

Exel Logistics

Freddie Mac

GE Capital Real Estate

Government of Singapore Investment

Corporation (GSIC)

Heitman Capital Management Corporation

Kennedy-Wilson International

Management Reports, Inc. (MRJ)

*New for 2001

Legg Mason

Lend Lease Real Estate Investments

PricewaterhouseCoopers

Prudential Real Estate Investors

The Roulac Group

RREEF

MIG Realty Advisors

National Association of Real Estate Investment Trusts (NAREIT)

Mortgage Banker's Association (MBA)

National Association of Industrial and Office Properties (NAIOP)

National Association of REALTORS® (NAR)

National Investment Center for the Seniors

Housing & Care Industries (NIC)

National Multi Housing Council (NMHC)

New York University Real Estate Institute

Real Estate Center at Texas A&M University

Realty One

Research Institute for Housing America

Society of Industrial and Office REALTORS® (SIOR)

South-Western Thomson Learning*

SSR Realty Advisors

Steven L. Newman Real Estate Institute

Torto Wheaton Research

UBS Brinson Realty Investors

Urban Land Institute (UU)

THE AMERICAN REAL ESTATE SOCIETY

2001 President Joseph B. Lipscomb Texas Christian University

Executive Director Theron R. Nelson* University of North Dakota

Associate Executive Directors M. Tim Allen Florida Atlantic University Roy T. Black Georgia State University Glenn E. Crellin Washington State University

Editor, JRER KoWang California State University - Fullerton

Editor, JREL Karl L. Gunterman Arizona State University Co-Editor, JREPM Willard McIntosh * Prudential Real Estate Investors Editor, JREPE Marvin Wolverton Washington State University Newsletter Editor Stephen A. Pyhrr Kennedy- Wilson International Doctoral Seminar Director Mark Dotzour Texas A&M University

Michael A. Ainikeeff (1998-02) Johns Hopkins University James R. DeLisle* (2002-06) Georgia State University Geoffrey Dohrmann (2000-04) Institutional Real Estate, Inc. Richard B. Gold (1999-03) Boston Financial Jacques Gordon (2001-05) LaSalle Investment Management

Karl L. Guntermann* (2000-04) Arizona State University Jun Han (1999-03) John Hancock Real Estate Investment

Group

John Baen 2000-02 University of North Texas

2001 Membership

Officers President Elect Christopher Manning Loyola Ma'}'mount University

Secretarytrreasurer Karl L. Guntermann Arizona State University

Mark G. Dotzour Texas A&M University

Arthur Margon New York University Joseph L. Pagliari, Jr. Citadel Realty

Elections Officer William C. Goolsby University of Arkansas-Little Rock

Director of Publications G. Donald Jud* University of North Carolina

Greensboro Co-Editor, JREPM Marc Louargand Cornerstone Real Estate Advisors Inc. Managing Editor, JREPE William G. Hardin III Mississippi State University Meeting Planner Arthur L. Schwartz, Jr.' University of South Florida

Parliamentarian Joseph D. Albert' James Madison University

Board of Directors G. Donald Jud* (2001-05) University of North Carolina-

Greensboro Ronald W. Kaiser (2002-06) Ballard. Biehl & Kaiser Marc A. Louargand (2000-04) Cornerstone Real Estate Advisors Inc. Willard Mclntosh* (1998-02) AIG Global Real Estate Investment

Corporation Issac Megbolugbe (2002-06) Fannie Mae Foundation Norman G. Miller (2001-05) University of Cincinnati Glenn R. Mueller* (1999-03) Johns Hopkins University &

Legg Mason

IRES Board Members M. Atef Sharkawy 2001-03 Texas A&M University

Vice President and Program Chair

Youguo Liang Prudential Real Estate Investors

Director of Development James R. Webb* Cleveland State University Director, Strategy Stephen E. Roulac* The Roulac Group &

University of Ulster Historian Walt A. Nelson Southwe.vt Missouri State University

ARES Webmaster Michael S. Young RREEF

Master of Ceremonies Stephen E. Roulac* The Roulac Group &

University of Ulster Co-Editor, JREPM Glenn R. Mueller fohns Hopkins University &

Legg Mason

Director of International Liaison Gracme Newell University afWestern Sydney

Ombudsperson Larry E. Wofford* C&L Systems

Richard Marchitelli (1998-02) Appraisal Institute Mauricio Rodriguez (2002-06) Texas Christian University Arthur L. Schwartz, Jr* (2001-05) University of South Florida Grant Thrall (1999-03) University of Florida John Williams (2000-04) Marghouse College Elaine M. Worzala (1998-02) Colorado State University

James R. Webb 2002-04 Cleveland State University

Academic ($110), Professional ($225), Academic Library ($350), Company Library ($450), Student ($55), Corporate ($450), Sponsor ($1,500), Regent ($3,000) and President's Council ($6,000). Checks should be made payable to ARES and correspondence should be addressed to: Helen Murphy, ARES, University of North Dakota. College of Business & Public Administration, P. O. Box 7120, Grand Forks, NO 58202-7120. Phone: 701-777-3670; Fax: 701-777-6380. Website: www.ARESnet.orgor Helen Murphy at helen_murphy@und.nodak.edu.

*Past President

2001 Rea I

Fellows of the American Estate Society

Endowed Doctoral Sponsorships

Glenn R. and Jan H. Mueller Theron R. and Susan L. Nelson James R. and Anais B. Webb

Joseph D. Albert James Madison University

Brent W. Ambrose University of Kentucky

Randy 1. Anderson Prudential Real Estate Investors'

Michael A. Anikeeff Johns Hopkins University

John S. Baen University of North Texas

John D. Benjamin American University

Donald H. Bleich California State University - Northridge

Amy Bogdon Fannie Mae Foundation

Waldo L. Born Eastern Illinois University

James H. Boykin Virginia Commonwealth University

Nicholas Buss PNC Bank'

Todd A. Canter ABKB I LaSalle Securities

James Carr Fannie Mae Foundation

Lijian Chen Lend Lease Real Estate Investments

Ping Cheng Salisbury State University

Mark S. Coleman Bondspace

James R. Cooper Georgia State University

Glenn E. Crellin Washington State University

Charles G. Dannis Crosson Dannis

Karen G. Davidson Davidson & Associates

James R. DeLisle Georgia State University

Gene Dilmore Realty Researchers

Geoffrey Dohrmann Institutional Real Estate Inc.

Mark G. Dotzour Texas A & M University

John T. Emery Louisiana Tech University

Fellows

Donald R. Epley Washington State University

Robert A. Ernst RAE Securities

Jack P. Friedman Jack P. Friedman & Associates'

S. Michael Giliberto J. P. Morgan Investment Management

John L. Glascock George Washington University

Paul R. Goebel Texas Tech University

Richard B. Gold Lend Lease Real Estate Investments

William C. Goolsby University of Arkansas - Little Rock

Jacques Gordon LaSalle Investment Management

G. Hayden Green University of Alaska - Anchorage

D. Wylie Greig RREEF

Karl L. Guntermann Arizona State University

Otis E. Hackett Otis E. Hackett & Associates

Thomas Hamilton Indiana State University

Jun Han John Hancock Real Estate Investments Group

Richard L. Haney Texas A & M University

William G. Hardin, III Mississippi State University

William T. Hughes MIG Realty Advisors

Jerome R. Jakubovitz MAl

G. Donald Jud University of North Carolina - Greensboro

Ronald W. Kaiser Ballard, Biehl & Kaiser*

Steven D. Kapplin University of South Florida

George R. Karvel University of Saint Thomas

William N. Kinnard. Jr. Real Estate Counseling Group of Connecticut

Richard Knitter Great Realty Advisors

2001 Fellows (continued)

Phillip T. Kolbe University of Memphis

Steven P. Laposa Pri ce water h 0 u se C 00 per s

Youguo Liang Prudential Real Estate Investors

Frederick Lieblich SSR Realty Advisors

Joseph B. Lipscomb Texas Christian University

Marc A. Louargand Cornerstone Realty Advisers Inc.

Emil Malizia University of North Carolina - Chapel Hill

Christopher A. Manning Loyola Marymount University

Richard Marchilelli Appraisal Institute

John F. McDonald University of Illinois - Chicago Circle

Willard McIntosh AIG Global Real Estate Investment Corporation

Isaac Megbolugbe Fannie Mae Foundation

Ivan 1. Miestchovich, Jr. University of New Orleans

Norman G. Miller University of Cincinnati

William Mundy Mundy Jarvis & Associates

F. C. Neil Myer Cleveland State University

Graeme Newell University of Western Sydney

Joseph L. Pagliari, Jr. Citadel Realty

Joseph D. Pasquarella Joseph D. Pasquarella & Co.

Edward F. Pierzak Henderson Investors North America

Steven A. Pyhrr Kennedy-Wilson International

R. Malcolm Richards Texas A&M University

Rudy R. Robinson, III Austin Valuation Consultants*

Stephen E. Roulac The Roulac Group & University of Ulster

Ronald C. Rutherford University of Texas-San Antonio

Anthony B. Sanders Ohio State University'

Karl-Werner Schulte European Business School

Arthur L. Schwartz, Jr. University of South Florida

David Scribner Scribner & Associates

M. Alef Sharkawy Texas A&M University

Leon Shilton Fordham University

Robert A. Simons Cleveland State University

Pelros Sivitanides Torto Wheaton Research

C. Ray Smith University of Virginia

Simon A. Stevenson University College-Dublin*

Stephen F. Thode Lehigh University

Robert Thompson King Sturge

Grant I. Thrall University of Florida

Raymond Torto Torto Wheaton Research

Raymond Y. C. Tsc Hong Kong Polytechnic

Jorge I. Vallejo Vallejo & Vallejo

Stephen M. Verba Realty One

Ko Wang California State University-Fullerton

R. Bryan Webb UBS Brinson Realty Advisors

John E. Williams Morehouse College

Larry E. Wofford C & L Systems Corporation

Marvin Wolverton University of Nevada-Las Vegas

Elaine M. WorzaIa Colorado State University

Charles H. Wurtzebach Henderson Properties

Tyler Yang Freddie Mac

Michael S. Young RREEF

Leonard V. Zumpano University of Alabama

*New for 200 1

Contents

About the Editors

Foreword

Introduction

Section I: Appraiser Decision Making and Valuation Accuracy

1. Behavioral Research into the Real Estate Valuation Process: Progress Toward a Descriptive Model Julian Diaz III / J. Andrew Hansz

2. Are Appraisers Statisticians? R. Kelley Pace / c.F. Sirmans / V Carlos Slawson, Jr.

3. The Components of Appraisal Accuracy Paul Gallimore

Section II: Regression, Minimum-Variance Grid Method, and Other Valuation Modeling Techniques

4. An Investigation of Property Price Studies Kicki Bjorklund / Bo Soderberg / Mats Wilhelmsson

5. Comparison of the Accuracy of the Minimum-Variance Grid Method and the Least Squares Method - a Non-Linear Extension Kuong Wing Chau / Wei Huang / Fung Fai Ng / Hin Man Louise Ng-Mak

xiii

xv

xvii

3

31

45

63

95

x

6. Error Trade-offs in Regression Appraisal Methods Max Kummerow / Hanga Galfalvy

7. Automated Valuation Models R. Kelley Pace / c.p Sirmans / V Carlos Slawson, Jr.

8. A Note on the Hedonic Model Specification for Income Properties Bo Soderberg

9. Neural Network vs. Hedonic Price Model: Appraisal of High-Density Condominiums K. C. Wong / Albert T.P. So / Y. C. Hung

Section III: Appraising Contaminated Property

10. Comparative Studies of United States, United Kingdom and New Zealand Appraisal Practice: Valuing Contaminated

105

133

157

181

Commercial Real Estate 201 William N. Kinnard, Jr. / Elaine M. Worzala / Sandy G. Bond / Paul J. Kennedy

11. Hedonic Modeling in Real Estate Appraisal: The Case of Environment Damages Assessment 227 Alan K. Reichert

12. Do Market Perceptions Affect Market Prices? A Case Study of a Remediated Contaminated Site 285 Sandy G. Bond

Section IV: Property Tax Assessment

13. Valuation of Land Using Regression Analysis 325 Mark A. Sunderman / John W. Birch

14. Grid-Adjustment Approach - Modem Appraisal Technique 341 Shwu-huei Huang

xi

Section V: New Perspectives on Traditional Appraisal Methods

15. The Unit-comparison Cost Approach in Residential Appraisal 357 Peter F. Colwell I David W Marshall

16. The Long-run Equilibrium Relationship among Equity Capitalization Rates for Retail, Apartment, Office, and Industrial Real Estate 373 Michael Devaney

17. A Fuzzy Discounted Cash Flow Analysis for Real Estate Investment 389 Tien Foo Sing I David Kim Hin Ho I Danny Poh Huat Tay

18. Real Options and Real Estate: A Review and Valuation Illustration 411 Steven H. Ott

Index 425

About the Editors

KoWang

Ko Wang is currently a professor in the Department of Finance and co­director of the Real Estate and Land Use Institute at the California State University - Fullerton. Prior to this appointment, Professor Wang was an assistant professor in the Department of Finance at the University of Texas at Austin and a chaired real estate professor at The Chinese University of Hong Kong. He had also worked full time for several leading real estate consulting and development firms in the U.S. and Asia. Professor Wang holds an MS degree in Community and Regional Planning (1982), an MBA degree with a real estate concentration (1984), and a Ph.D. degree in real estate and finance (1988). All the degrees are from the University of Texas at Austin.

Professor Wang publishes both in real estate and finance journals, which include Real Estate Economics and Journal of Real Estate Research (10 pub­lications each), Journal of Urban Economics, Journal of Real Estate Finance and Economics, International Real Estate Review, Journal of Finance, Journal of Financial Economics, Journal of Financial and Quantitative Analysis, and Review of Financial Studies. He also co-authored a book on Real Estate Investment Trusts (Oxford University Press, New York) and co-wrote 18 teaching cases and teaching notes on real estate and finance issues faced by firms in Asia (distributed by Harvard Business School Publishing and European Case Clearing House). He is the editor of the Journal of Real Estate Research and the founding editor of the International Real Estate Review. Professor Wang is a Fellow of the Homer Hoyt Institute and the Hong Kong Institute of Real Estate.

XIV

Marvin L. Wolverton

Marvin L. Wolverton is currently a visiting associate professor in the Department of Finance at the University of Nevada Las Vegas. He holds a Ph.D. in Business Administration with Real Estate and Decision Science con­centrations from Georgia State University in Atlanta, Georgia, and he holds the Appraisal Institute's MAl designation. He also has a M.S. in Economics from Arizona State University in Tempe, Arizona, and a BSEM in Mining Engineering from the New Mexico Institute of Mining and Technology in Socorro, New Mexico. Professor Wolverton taught in the Department of Finance, Insurance, and Real Estate at Washington State University for five years prior to his present affiliation with UNLV, where he held the positions of Alvin 1. Wolff Professor of Real Estate and Director of Real Estate Research. Prior to his university affiliations, he worked for many years as a practicing real estate appraiser and consultant, and continues to engage in consulting assignments as time permits.

Dr. Wolverton is on the editorial boards of The Appraisal Journal and the Journal of Real Estate Research. He is also the editor of the Journal of Real Estate Practice and Education. He has written numerous articles appearing in The Appraisal Journal, the Journal of Real Estate Research, Real Estate Economics, the Journal of Real Estate Finance and Economics, the Assessment Journal, the Journal of Property Research, the Journal of Property Investment and Finance, and Real Estate Review. He and his wife Mimi now reside in Las Vegas, Nevada.

Foreword

The Appraisal Institute is pleased to join with the American Real Estate Society in cosponsoring this monograph entitled "Real Estate Valuation Theory." Since its inception the Appraisal Institute has strongly supported research and theoretical thought as a means to expand our body of knowledge. In that regard we have long supported the efforts and undertakings of the American Real Estate Society.

As with most fields of research and investigation, the ultimate goal is to refine our modes of practice. Those of us who work on the frontlines of val­uation practice look to theoreticians to discover new ways of thinking, new approaches and new applications that change and improve the way we solve our clients' problems. Theory, combined with technology, has created a new dimension for us to work in, which is pointed up in the range of articles presented in this monograph.

The Appraisal Institute appreciates the opportunity to participate in the publication of "Real Estate Valuation Theory," and we look forward to con­tinuing our close relationship with the American Real Estate Society and its members in the future.

Brian A. Glanville, MAl 2001 President

Appraisal Institute

Introduction

Is real property appraisal evolving, or have all of the "big ideas" already been conceived? This question has been circulating among the leaders of the Appraisal Institute during the past several annual meetings, usually encoun­tering little in the way of satisfying answers. This isn't too surprising, because grand ideas and new knowledge don't usually evolve from off-hand con­versation. They are the product of focused intellectual activity and hours of difficult work.

This unquenched thirst for new knowledge is the primary reason for assembling this collection of new manuscripts dealing with valuation theory, which was financially underwritten by the Appraisal Institute. Their generos­ity and willingness to partner with the American Real Estate Society made this collection of thoughtful and thought provoking essays possible. They are the result of a global response to a worldwide call for papers, and demonstrate that real estate valuation is indeed an international discipline. The United States, Australia and New Zealand, Southeast Asia, the Pacific Rim, and Europe are all represented by this impressive collection of authors. Together, the eighteen essays that make up this volume demonstrate that there are a sufficient number of "big ideas" to challenge and improve the appraisal profession for years to come.

The monograph is organized around five categories of intellectual contri­bution to the whole - appraiser decision making and valuation accuracy, application of nontraditional appraisal techniques such as regression and the minimum-variance grid method, appraising contaminated property, ad val­orem tax assessment, and new perspectives on traditional appraisal methods. One common thread is that all of the papers are exceptionally well written and thought provoking.

XVlll

Section I: Appraiser Decision Making and Valuation Accuracy

This section begins with a paper written by Julian Diaz III and J. Andrew Hansz titled "Behavioral Research into the Real Estate Valuation Process: Progress Toward a Descriptive Model." The authors rightly claim that a descrip­tive model must be derived to fully understand expert appraiser problem­solving and decision-making behavior, hence the appraisal process. They set the scene for this section and for the monograph by stressing the human element of appraisal. Fortunately for the appraisal discipline, much of the behavioral research done in the field of real estate has been focused on appraisers. Their paper reviews this research, most of which deals with com­mon cognitive biasing influences on appraisers. They discuss the underlying behavioral theories, organize extant research findings and the techniques employed, and take a small but crucial initial step toward development of a descriptive model of the valuation process. In their descriptive model the appraisal process varies with appraiser experience, market familiarity, the reg­ulatory environment, client relationships, and the cultural context of the work environment.

Section I's second paper is titled "Are Appraisers Statisticians." Here R. Kelley Pace, C.P. Sirmans, and V. Carlos Slawson, Jr. look at the phenom­enon of "statistically challenged [human] appraisers following ad hoc proce­dures" oftentimes exhibiting prediction errors substantially less than those exhibited by hedonic pricing models resting on a foundation of "statistical and economic theory." They note that rather than relying on the human equiv­alent of a statistical hedonic model, appraisers seem to focus on the compa­rable sale selection process. They investigate a spatial-temporal model based solely upon distance and time and a multi-dimensional model adding baths, bedrooms, area, and age as exemplars of what appraisers do (focus on com­parable sales), and compare the results of these models to a simple form and a complex form of hedonic model. The two models designed to capture what appraisers do both outperformed the two hedonic pricing models derived for comparison purposes. This leads the authors to the conclusion that "the statistically naIve practice of appraisers may have a more sophisticated basis than a casual examination would reveal." They suggest that prediction in statistical models could be improved by incorporation of elements of what appraisers actually do.

Paul Gallimore completes this section with his paper, "The Components of Appraisal Accuracy." He points out the correspondence between the concepts of appraisal accuracy (comparison of appraised values and market values) and appraisal variance (agreement among appraisers valuing the same property) and the research design conditions of validity and reliability. He

XIX

presents a framework - the lens model - that provides insight into elements of appraisal accuracy. The lens model's strength is its applicability to problem solving by use of cues (i.e., comparable sales) to make judgments about events that cannot be observed directly (i.e., property value). This theoretical framework, borrowed from psychology, uncovers three avenues for improv­ing appraisal accuracy - predictability of the real property environment, the match between valuation models and the real property environment, and con­sistency in application of valuation models. Gallimore concludes that appraisal accuracy can be improved by recognizing appraisers as being "for­casters of unobservable events." He then suggests that this perspective leads to the realization that improvements in data, monitoring of external influ­ences, and greater consistency in application of valuation models may all be fruitful means to the goal of improving appraisal accuracy.

Section II: Regression, Minimum-Variance Grid Method, and Other Valuation Modeling Techniques

Kicki Bjorklund, Bo SOderberg and Mats Wilhelmsson's massive under­taking, "An Investigation of Property Price Studies," appropriately leads off this section of the monograph. They systematically examine 145 articles appearing in 13 high ranking American and European real estate, housing, and urban economics journals during the 1990--1995 period, assessing how well the researchers adhered to "general methodological guidelines associated with the treatment and reporting of empirical econometric research." They also provide an overview of real estate economics topics covered in these papers and classify the articles into subcategories.

The 145-article sample represents distillation of 1,882 articles, leaving only those manuscripts incorporating estimates of price or rent equations using regression techniques. The 145 articles are assessed from four perspec­tives, including the relation to previous research, modeling procedures, the data, and how the results are presented and interpreted. Their analysis reveals that there is room for improvement in hedonic modeling and reporting of research results. Two particularly troubling findings are (i) residual analysis, a fundamental element of regression modeling, was omitted from 75% of the articles studied, and (ii) motivation for choice of functional form was omitted from 73% of the articles. Clearly, real estate and housing practitioners and researchers can improve on reporting of their work, and the 18-item checklist found in this paper represents a good source of guidance.

The second article in this section is "Comparison of the Accuracy of the Minimum-Variance Grid Method and the Least Squares Method - a Non­Linear Extension" by Kwong Wing Chau, Wei Huang, Fung Fai Ng, and Hin

xx

Man Louise Ng-Mak. They take on the question of comparing the accuracy of the least squares method to the relatively new minimum-variance grid method, extending the work in this area of inquiry to include the non-linear least squares estimator case. Importantly, these authors derive the necessary and suf­ficient conditions under which erG < erNLS (conditions where the minimum­variance grid method is more accurate than the non-linear least squares method based on prediction accuracy). They note, however, that the results for the non­linear case are asymptotic, requiring a large number of observations, which limits application to liquid markets with readily available market data. They conclude by calling for empirical tests of their mathematical result.

In "Error Tradeoffs in Regression Appraisal Methods," Max Kummerow and Hanga Galfalvy conduct an empirical study of an issue related to the pre­ceding paper by Chau, et ai., and the Pace, et ai., paper from Section I. That is, do market sales data become so heterogeneous as sample size grows that the errors stemming from such heterogeneity more than offset the asymptotic benefits derived from a large sample? They argue that one would expect measurement errors and omitted variables to become more problematic as sample size grows, perhaps leading to greater prediction error in large sam­ples. They demonstrate error tradeoffs between decreasing random error and growing adjustment errors as sample size increases, and test their models using data consisting of more than 2,000 sales derived from Perth, Australia. Kummerrow and Galfalfy postulate a "U" shaped prediction error function based on analysis of their data, consistent with accepted industry practice of basing an analysis on a small number of most comparable sales.

In this section's fourth article, "Automated Valuation Models," R. Kelley Pace, c.F. Sirmans, and V. Carlos Slawson, Jr. examine the prospect of adapt­ing computer aided mass assessment (CAMA) used in property tax assessment to traditional appraisal settings. They contrast the accuracy of manual and CAMA appraisals, discuss issues of moral hazard, unbundling appraisal serv­ices, the market for low-cost, automated appraisals, and the possibility of inte­grating traditional and automated appraisal services. The article then explores implementation of an automated valuation system, dealing with a variety of related issues such as data needs, omitted variables, model construction, model diagnostics, and the spatial and temporal nature of real estate data. They note also that the last item - dealing with spatial and temporal information -constitutes the major computational problem for adapting CAMA to traditional appraisal work. The authors conclude that an ideal system in terms of compet­itive pricing combined with objective and acceptably accurate results might involve appraisers as property examiners and data collectors along with CAMA techniques being employed to estimate property values.

Bo Soderberg narrows the focus of the hedonic modeling discussion to income property in this section's fifth paper, "A Note on the Hedonic Model

xxi

Specification for Income Properties." The study is developed in the context of Stockholm, Sweden where a mixed-use assumption consisting of residential and commercial property components underlies the analysis. Using a 282-observation sample, the author compares results derived from a Cobb­Douglas type, log-linear value estimation model (Model I) with results of a semi-log linear estimation model resulting from underlying exponential rela­tionships (Model II). The estimation equations include variables controlling for proportions of residential and commercial space, building age, distance from the central business district, nature of the buyer, and temporal variation over 12 quarters. Although regression results were nearly identical in terms of goodness of fit, age and distance effects vary substantially between the two models. There was also a difference in out-of-sample prediction accuracy, with Model I being inferior to Model II in these measures. Appendices to the article provide valuable theoretical insight into the observed superior per­formance of the Model II functional form.

K.c. Wong, Albert T.P. So, and Y. C. Hung wrote the final paper in this section. In "Neural Network vs. Hedonic Price Model: Appraisal of High­Density Condominiums," they select a sample of 216 residential condo­minium units in Hong Kong as a training set to value 35 units held out as a test set. These data are used to compare the results of a neural network with two alternative specifications of the hedonic pricing model. They find that the neural network may appeal to a "risk averse appraiser" wanting to reduce maximum error on individual property appraisals, although the neural net­work did not significantly reduce average prediction error. They note the drawbacks to neural networks, including long computation time and the absence of indicators of implicit prices of property characteristics. Conversely, in addition to underscoring the reliability of, and potential for, neural networks as a valuation tool, their paper provides important insight regarding neural network design decisions concerning the number of itera­tions and hidden nodes.

Section III: Appraising Contaminated Property

Valuing contaminated property and estimating real property damages resulting from contamination represent a relatively new, yet substantial, aspect of modern appraisal practice. According to William N. Kinnard, Jr., Elaine M. Worzala, Sandy Bond, and Paul J. Kennedy, in "Comparative Studies of United States, United Kingdom and New Zealand Appraisal Practice: Valuing Contaminated Commercial Property," U.S. literature on this topic began recently in approximately 1984. These authors trace this literature in order to compare how appraisers have dealt with contamination related

xxii

valuation assignments in the U.S., U.K. and New Zealand. They also report on surveys of appraisers in the three countries (and Canada) in order to ana­lyze the extent to which practice adheres to techniques recommended in pub­lished articles and conference papers. Lastly, they review practice standards relating to valuing contaminated property in light of their research findings regarding practice and the literature. Their research reveals dramatic between­country differences in contaminated property valuation experience and prac­tice. Additionally, data availability varies widely by country. These two factors lead to divergent opinions and little agreement regarding best prac­tices. While this lack of agreement may be unsettling, it underscores the need to develop better methods for dealing with contaminated property. It also clearly identifies a need to build a better understanding of the analysis tools currently being used. The remaining two articles in this section are a first step toward meeting this need.

"Hedonic Modeling in Real Estate Appraisal: The Case of Environment Damages Assessment," by Alan K. Reichert is a thorough primer on applying the hedonic modeling technique to contamination appraisal assignments. The paper confronts practical problems that must be addressed "to ensure accurate and reliable results." The issues discussed include variable selection, func­tional form, and sample size effects-viewed from multiple perspectives including goodness of fit, prediction error, number of significant variables, and average independent variable significance level. Also discussed are con­trol sample selection and double-counting problems that can arise from over­modeling. Three differing data sources (MLS, PACE and METROSCAN) are evaluated on the basis of data limitations and prediction accuracy in the con­text of an actual valuation case, and accuracy of longitude and latitude coor­dinates derived from mapping software is assessed. Other important and insightful topics include the interaction of multicollinearity and sample size, heteroskedasticity, and interpretation of regression coefficients. The author concludes that an appropriately designed and estimated hedonic model can produce median absolute percentage errors small enough to compete "with maximum error rates achieved by appraisers and tax assessors."

In this section's last paper, "Do Market Perceptions Affect Market Prices? A Case Study of a Remediated Contaminated Site," Sandy Bond compares perceptions uncovered by survey to the structure of a hedonic pricing model. The study is focused on post-remediated vacant residential land in Perth, Australia. The opinion survey questionnaire consisted of 50 response items developed specifically for the research project. The survey results revealed perceptions and attitudes concerning location quality, river view, access to public transportation, and proximity to a remediated site with contaminants buried on-site. The hedonic model was employed in the study to investigate the degree to which the perceptions and attitudes uncovered by the survey

XXlll

actually impacted upon market prices. The opinion survey results were shown to be "at least consistent with those of the market sales analysis," showing similar directions of impact. One important finding was that the stigma asso­ciated with a remediated site's contamination history can be offset by location and site amenities. Another insight to be gained from this article is the impor­tance of fully understanding prevailing perceptions and attitudes and identi­fying variables that capture these for use in a price estimation model.

Section IV: Property Tax Assessment

Mark A. Sunderman and John W. Birch take on the issue of estimating land value when there are a limited number of available land sales in "Valuation of Land Using Regression Analysis." Noting that improved prop­erty sale price is composed of land value and improvement value, they reason that a multiple regression model can be used to isolate and estimate the value of the land, improvements, or both. For improved property, land value is esti­mated by deducting market-derived estimates of improvement value from the estimated market value of the total property (land and improvements). Land sales are included in the model along with improved sales by setting all land­sale improvement characteristic variables equal to zero. Their model also cre­atively accounts for neighborhood effects such as systematic price-size nonlinearity within a given neighborhood and differences in price levels between neighborhoods. Based on a holdout sample, the model derive by the authors was more accurate at predicting land price than the conventional sys­tem in use at the time of the study. The practicality of this paper is enhanced by a thorough description of how the authors adapted their regression results to an existing system of neighborhood land value tables, which was restricted to six different lot sizes. The model, employed in an assessment context here, is applicable to land value appraisals in other contexts as well.

Shwu-huei Huang contributes this section's other paper, "Grid­Adjustment Approach - Modem Appraisal Technique." The paper describes inefficiencies and inequities in the existing land declared price system of esti­mating the value of the tax base for Taiwan's land value tax, and then pro­poses two methods for improving the system. Inefficiency is evident in current assessment-to-sale-price ratios (A-S Ratio) averaging from.1O to.13 citywide over the study period. Inequity is revealed in the form of unaccept­ably high coefficients of variation and failure of a progressivity test. One pro­posed cure is a hedonic model, derived after extensive experimentation with functional form and heteroskedasticity correction. This method improved citywide A-S Ratios to an average of 1.08 over the five-year study period, reduced the average coefficient of variation to within 10% of Back's criterion,

XXIV

and removed all evidence of progressivity or regressivity. The grid-adjustment approach is the second proposed cure. The results were similar to the hedonic modeling approach, with improvements in the three metrics used to measure efficiency and equity. The author recommends the grid-adjustment approach for the practical reason that assessors are more likely to be accepting of it due to similarities with the familiar market comparison approach.

Section V: New Perspectives on Traditional Appraisal Methods

In "The Unit-Comparison Approach in Residential Appraisal" Peter F. Colwell and David W. Marshall develop a non-tabular approach to construc­tion cost estimation. They derive a three-component function that incorpo­rates constant, linear, and square root elements. Constant elements are those that are found in a base house regardless of size. Examples include a mini­mum number of plumbing fixtures, a front door, and elements of the heating and cooling systems. Linear elements are those that vary directly with house size such as floors and ceilings. Square root elements include components that increase in cost with size at a decreasing rate such as exterior walls. The cost function developed in their paper is the sum of these three components, and is shown to be quite adaptable. The cost estimating equation developed in this paper has potential to be used in practice to evaluate the accuracy of cost man­uals. Another application might be to replace cost tables with cost functions imbedded into a computerized cost program. Additionally, as the authors note, the results here have theoretical implications regarding the expected concav­ity of hedonic functions in square footage. Finally, in an appendix to the paper, the authors develop additional and interesting cost details, including separation of the heating and cooling equipment from the air circulation sys­tem and incorporation of roof pitch variation into the cost function.

Michael Devaney examines integration of equity capitalization rates in "The Long-Run Equilibrium Relationship among Equity Capitalization Rates for Retail, Apartment, Office, and Industrial Real Estate." He utilizes a vector error correction model to investigate if and how equity capitalization rates converge toward a long-run equilibrium. Equity capitalization rates are derived from American Council of Life Insurance overall rate data spanning a 23-year period using the band of investment technique. Results indicate that retail, apartment, and industrial equity capitalization rates exhibit a long-run equilibrium relationship attained primarily through adjustments in apartment and retail pricing. Office equity capitalization rates were, on the other hand, segmented from rates determined in the other three property type markets.

xxv

The office anomaly could have been a consequence of market imbalance in office properties during the 1980s and early 1990s, as the author notes, and may be less apparent during more normal periods. The findings in the retail, apartment, and industrial markets is consistent with the principle of substitu­tion, which suggests that mispricing of a given property type will be corrected by market activity. The paper's empirical evidence of cointegration has impli­cations for appraisers and investors confronting markets demonstrating devi­ations from their long-run relationships.

In "A Fuzzy Discounted Cash Flow Analysis for Real Estate Investment," Tien Foo Sing, David Kim Hin Ho, and Danny Poh Hat Tay explore applica­tion of fuzzy set theory to real property cash flow modeling. As one of the few papers dealing with application of fuzzy mathematics to financial modeling, this paper represents a first step into uncharted waters for the real property valuation discipline. The authors provide a case illustration as a means of demonstrating the insight gained by fuzzy discounted cash flow analysis compared to the traditional, deterministic discounted cash flow (DCF) tech­nique. They guide the reader through unfamiliar fuzzy mathematical concepts, fuzzy financial modeling, and the development of a fuzzy DCF model. This new knowledge is then applied to an office development case in Singapore as a means of comparing fuzzy DCF and traditional DCF results. They conclude that the fuzzy DCF model "provides a natural and intuitive way of dealing with cognitive uncertainty associated [with] vague and impre­cise information."

The final paper in this section is "Real Options and Real Estate: A Review and Valuation Illustration" by Steven H. Ott. The author notes that application of the discounted cash flow "NPV decision rule ignores the changing dynam­ics of the actual marketplace." Specifically, traditional DCF models ignore a property owner's inherent flexibility to make future choices in order to adapt to changing market conditions. Real option analysis takes into account future investment opportunities imbedded in real estate investment decisions. Incorporation of real options analysis into analytical models can enhance real estate decision-making and value estimation. This paper provides a practical illustration of how this can be done using an actual Charlotte, North Carolina land development project, linking the value of a growth option to the value derived through the traditional DCF method. The imbedded growth option turns what would be traditionally viewed as an unfeasible, negative NPV proj­ect into a project viewed as a feasible, positive NPV project. This paper pro­vides new insight into the highest and best use issue, which appraisers frequently encounter. Indeed, when option value is included in overall real property value, it may become more difficult to justify a change in an existing use.

XXVI

Conclusion

The editors thank the Appraisal Institute, the American Real Estate Society and all of the contributing authors and referees for their combined and indi­vidual efforts, without which this volume would have never existed. We also acknowledge the efforts of Wayne Archer for handling 3 of the papers that ultimately were accepted for publication here. We are pleased to have had the opportunity to edit this volume, and hope that it proves to be a useful resource for years to come.